One Health Lens for Antimicrobial Resistance Research and Funding: A Systematic Review.

One Health (OH) offers conceptual and applied prospects to advance planetary health and integrative biology in the 21st century. For example, The World Health Organization (WHO) has declared antimicrobial resistance (AMR) one of humanity's top 10 health threats worldwide (AMR). The AMR research, as seen through the OH lens, recognizes the interdependence and the coproduction of the health of humans, nonhuman animals, and the environment (the OH triad). Moreover, research and development (R&D) is required to generate potential solutions to prevent, diagnose, and treat infections and control the spread and emergence of AMR. However, it is still unclear how well the OH approach is integrated into current AMR R&D. In this study, we present a systematic review on the OH funding landscape for cross-sectoral AMR R&D, and its alignment/gaps with the current global strategic agenda on AMR. A systematic literature review was conducted using public databases covering the period between January 2015 and May 2022. We included the studies and reviews on AMR encompassing more than one sector of the OH triad. Out of the 777 included studies, 475 (61%) encompassed the three OH sectors. A key finding of the present systematic review is that the environment was the most neglected sector in the OH triad. AMR surveillance, transmission, and interventions are the most commonly studied priority topics. In addition, both cross-sectoral AMR literature and investments have been increasing since 2017. The operational aspect of AMR is the most researched and funded area. However, certain priority topics in the strategic research and innovation agenda of the Joint Programming Initiative on AMR are underrepresented in OH AMR research, such as diagnosis and therapeutics. To the best of authors' knowledge, this is the first study that systematically reviews the cross-sectoral literature on AMR, classifies it, and aligns and contextualizes it in regard to the funding landscape of AMR. This systematic review identifies neglected areas in AMR R&D and could serve as critical information for policymaking so as to realize the objectives of the Global Action Plan on AMR. Going forward, more cross-sectoral AMR research and funding are needed. As integrative biology and omics systems science are poised to benefit from a rapprochement with the OH lens, the present article highlights the AMR research and funding landscapes.

Host and microbiome features of secondary infections in lethal covid-19.

Secondary infections contribute significantly to covid-19 mortality but driving factors remain poorly understood. Autopsies of 20 covid-19 cases and 14 controls from the first pandemic wave complemented with microbial cultivation and RNA-seq from lung tissues enabled description of major organ pathologies and specification of secondary infections. Lethal covid-19 segregated into two main death causes with either dominant diffuse alveolar damage (DAD) or secondary pneumonias. The lung microbiome in covid-19 showed a reduced biodiversity and increased prototypical bacterial and fungal pathogens in cases of secondary pneumonias. RNA-seq distinctly mirrored death causes and stratified DAD cases into subgroups with differing cellular compositions identifying myeloid cells, macrophages and complement C1q as strong separating factors suggesting a pathophysiological link. Together with a prominent induction of inhibitory immune-checkpoints our study highlights profound alterations of the lung immunity in covid-19 wherein a reduced antimicrobial defense likely drives development of secondary infections on top of SARS-CoV-2 infection.

Proposal of a population wide genome-based testing for Covid-19.

Our lives (and deaths) have by now been dominated for two years by COVID-19, a pandemic that has caused hundreds of millions of disease cases, millions of deaths, trillions in economic costs, and major restrictions on our freedom. Here we suggest a novel tool for controlling the COVID-19 pandemic. The key element is a method for a population-scale PCR-based testing, applied on a systematic and repeated basis. For this we have developed a low cost, highly sensitive virus-genome-based test. Using Germany as an example, we demonstrate by using a mathematical model, how useful this strategy could have been in controlling the pandemic. We show using real-world examples how this might be implemented on a mass scale and discuss the feasibility of this approach.

A framework for validating AI in precision medicine: considerations from the European ITFoC consortium.

BACKGROUND: Artificial intelligence (AI) has the potential to transform our healthcare systems significantly. New AI technologies based on machine learning approaches should play a key role in clinical decision-making in the future. However, their implementation in health care settings remains limited, mostly due to a lack of robust validation procedures. There is a need to develop reliable assessment frameworks for the clinical validation of AI. We present here an approach for assessing AI for predicting treatment response in triple-negative breast cancer (TNBC), using real-world data and molecular -omics data from clinical data warehouses and biobanks. METHODS: The European "ITFoC (Information Technology for the Future Of Cancer)" consortium designed a framework for the clinical validation of AI technologies for predicting treatment response in oncology. RESULTS: This framework is based on seven key steps specifying: (1) the intended use of AI, (2) the target population, (3) the timing of AI evaluation, (4) the datasets used for evaluation, (5) the procedures used for ensuring data safety (including data quality, privacy and security), (6) the metrics used for measuring performance, and (7) the procedures used to ensure that the AI is explainable. This framework forms the basis of a validation platform that we are building for the "ITFoC Challenge". This community-wide competition will make it possible to assess and compare AI algorithms for predicting the response to TNBC treatments with external real-world datasets. CONCLUSIONS: The predictive performance and safety of AI technologies must be assessed in a robust, unbiased and transparent manner before their implementation in healthcare settings. We believe that the consideration of the ITFoC consortium will contribute to the safe transfer and implementation of AI in clinical settings, in the context of precision oncology and personalized care.

Global Open Health Data Cooperatives Cloud in an Era of COVID-19 and Planetary Health.

Big data in both the public domain and the health care industry are growing rapidly, for example, with broad availability of next-generation sequencing and large-scale phenomics datasets on patient-reported outcomes. In parallel, we are witnessing new research approaches that demand sharing of data for the benefit of planetary society. Health data cooperatives (HDCs) is one such approach, where health data are owned and governed collectively by citizens who take part in the HDCs. Data stored in HDCs should remain readily available for translation to public health practice but at the same time, governed in a critically informed manner to ensure data integrity, veracity, and privacy, to name a few pressing concerns. As a solution, we suggest that data generated from high-throughput omics research and phenomics can be stored in an open cloud platform so that researchers around the globe can share health data and work collaboratively. We describe here the Global Open Health Data Cooperatives Cloud (GOHDCC) as a proposed cloud platform-based model for the sharing of health data between different HDCCs around the globe. GOHDCC's main objective is to share health data on a global scale for robust and responsible global science, research, and development. GOHDCC is a citizen-oriented model cooperatively governed by citizens. The model essentially represents a global sharing platform that could benefit all stakeholders along the health care value chain.

Metagenomic identification, purification and characterisation of the Bifidobacterium adolescentis BgaC ß-galactosidase.

Members of the human gut microbiota use glycoside hydrolase (GH) enzymes, such as ß-galactosidases, to forage on host mucin glycans and dietary fibres. A human faecal metagenomic fosmid library was constructed and functionally screened to identify novel ß-galactosidases. Out of the 16,000 clones screened, 30 ß-galactosidase-positive clones were identified. The ß-galactosidase gene found in the majority of the clones was BAD_1582 from Bifidobacterium adolescentis, subsequently named bgaC. This gene was cloned with a hexahistidine tag, expressed in Escherichia coli and His-tagged-BgaC was purified using Ni2+-NTA affinity chromatography and size filtration. The enzyme had optimal activity at pH 7.0 and 37 °C, with a wide range of pH (4-10) and temperature (0-40 °C) stability. It required a divalent metal ion co-factor; maximum activity was detected with Mg2+, while Cu2+ and Mn2+ were inhibitory. Kinetic parameters were determined using ortho-nitrophenyl-ß-D-galactopyranoside (ONPG) and lactose substrates. BgaC had a Vmax of 107 µmol/min/mg and a Km of 2.5 mM for ONPG and a Vmax of 22 µmol/min/mg and a Km of 3.7 mM for lactose. It exhibited low product inhibition by galactose with a Ki of 116 mM and high tolerance for glucose (66% activity retained in presence of 700 mM glucose). In addition, BgaC possessed transglycosylation activity to produce galactooligosaccharides (GOS) from lactose, as determined by TLC and HPLC analysis. The enzymatic characteristics of B. adolescentis BgaC make it an ideal candidate for dairy industry applications and prebiotic manufacture.Key points• Bifidobacterium adolescentis BgaC ß-galactosidase was selected from human faecal metagenome.• BgaC possesses sought-after properties for biotechnology, e.g. low product inhibition.• BgaC has transglycosylation activity producing prebiotic oligosaccharides. Graphical Abstract.

Staphylococcus saccharolyticus: An Overlooked Human Skin Colonizer.

Coagulase-negative staphylococcal species constitute an important part of the human skin microbiota. In particular, facultative anaerobic species such as Staphylococcus epidermidis and Staphylococcus capitis can be found on the skin of virtually every human being. Here, we applied a culture-independent amplicon sequencing approach to identify staphylococcal species on the skin of healthy human individuals. While S. epidermidis and S. capitis were found as primary residents of back skin, surprisingly, the third most abundant member was Staphylococcus saccharolyticus, a relatively unstudied species. A search of skin metagenomic datasets detected sequences identical to the genome of S. saccharolyticus in diverse skin sites, including the back, forehead, and elbow pit. Although described as a slow-growing anaerobic species, a re-evaluation of its growth behavior showed that S. saccharolyticus can grow under oxic conditions, and, in particular, in a CO2-rich atmosphere. We argue here that S. saccharolyticus was largely overlooked in previous culture-dependent and -independent studies, due to its requirement for fastidious growth conditions and the lack of reference genome sequences, respectively. Future studies are needed to unravel the microbiology and host-interacting properties of S. saccharolyticus and its role as a prevalent skin colonizer.

Genomic and Ecogenomic Characterization of Proteus mirabilis Bacteriophages.

Proteus mirabilis often complicates the care of catheterized patients through the formation of crystalline biofilms which block urine flow. Bacteriophage therapy has been highlighted as a promising approach to control this problem, but relatively few phages infecting P. mirabilis have been characterized. Here we characterize five phages capable of infecting P. mirabilis, including those shown to reduce biofilm formation, and provide insights regarding the wider ecological and evolutionary relationships of these phages. Transmission electron microscopy (TEM) imaging of phages vB_PmiP_RS1pmA, vB_PmiP_RS1pmB, vB_PmiP_RS3pmA, and vB_PmiP_RS8pmA showed that all share morphologies characteristic of the Podoviridae family. The genome sequences of vB_PmiP_RS1pmA, vB_PmiP_RS1pmB, and vB_PmiP_RS3pmA showed these are species of the same phage differing only by point mutations, and are closely related to vB_PmiP_RS8pmA. Podophages characterized in this study were also found to share similarity in genome architecture and composition to other previously described P. mirabilis podophages (PM16 and PM75). In contrast, vB_PimP_RS51pmB showed morphology characteristic of the Myoviridae family, with no notable similarity to other phage genomes examined. Ecogenomic profiling of all phages revealed no association with human urinary tract viromes, but sequences similar to vB_PimP_RS51pmB were found within human gut, and human oral microbiomes. Investigation of wider host-phage evolutionary relationships through tetranucleotide profiling of phage genomes and bacterial chromosomes, indicated vB_PimP_RS51pmB has a relatively recent association with Morganella morganii and other non-Proteus members of the Morganellaceae family. Subsequent host range assays confirmed vB_PimP_RS51pmB can infect M. morganii.

The potential of nanoflow liquid chromatography-nano electrospray ionisation-mass spectrometry for global profiling the faecal metabolome.

Faeces are comprised of a wide array of metabolites arising from the circulatory system as well as the human microbiome. A global metabolite analysis (metabolomics) of faecal extracts offers the potential to uncover new compounds which may be indicative of the onset of bowel diseases such as colorectal cancer (CRC). To date, faecal metabolomics is still in its infancy and the compounds of low abundance present in faecal extracts poorly characterised. In this study, extracts of faeces from healthy subjects were profiled using a sensitive nanoflow-nanospray LC-MS platform which resulted in highly repeatable peak retention times (<2% CV) and intensities (<15% CV). Analysis of the extracts revealed wide coverage of the faecal metabolome including detection of low abundant signalling compounds such as sex steroids and eicosanoids, alongside highly abundant pharmaceuticals and tetrapyrrole metabolites. A small pilot study investigating differences in metabolomics profiles of faecal samples obtained from 7 CRC, 25 adenomatous polyp and 26 healthy groups revealed that secondary bile acids, conjugated androgens, eicosanoids, phospholipids and an unidentified haem metabolite were potential classes of metabolites that discriminated between the CRC and control sample groups. However, much larger follow up studies are needed to confirm which components of the faecal metabolome are associated with actual CRC disease rather than dietary influences. This study reveals the potential of nanospray-nanoflow LC-MS profiling of faecal samples from large scale cohort studies for uncovering the role of the faecal metabolome in colorectal disease formation.

Resolution of habitat-associated ecogenomic signatures in bacteriophage genomes and application to microbial source tracking.

Just as the expansion in genome sequencing has revealed and permitted the exploitation of phylogenetic signals embedded in bacterial genomes, the application of metagenomics has begun to provide similar insights at the ecosystem level for microbial communities. However, little is known regarding this aspect of bacteriophage associated with microbial ecosystems, and if phage encode discernible habitat-associated signals diagnostic of underlying microbiomes. Here we demonstrate that individual phage can encode clear habitat-related 'ecogenomic signatures', based on relative representation of phage-encoded gene homologues in metagenomic data sets. Furthermore, we show the ecogenomic signature encoded by the gut-associated ɸB124-14 can be used to segregate metagenomes according to environmental origin, and distinguish 'contaminated' environmental metagenomes (subject to simulated in silico human faecal pollution) from uncontaminated data sets. This indicates phage-encoded ecological signals likely possess sufficient discriminatory power for use in biotechnological applications, such as development of microbial source tracking tools for monitoring water quality.

Models of Models: A Translational Route for Cancer Treatment and Drug Development.

Every patient and every disease is different. Each patient therefore requires a personalized treatment approach. For technical reasons, a personalized approach is feasible for treatment strategies such as surgery, but not for drug-based therapy or drug development. The development of individual mechanistic models of the disease process in every patient offers the possibility of attaining truly personalized drug-based therapy and prevention. The concept of virtual clinical trials and the integrated use of in silico, in vitro, and in vivo models in preclinical development could lead to significant gains in efficiency and order of magnitude increases in the cost effectiveness of drug development and approval. We have developed mechanistic computational models of large-scale cellular signal transduction networks for prediction of drug effects and functional responses, based on patient-specific multi-level omics profiles. However, a major barrier to the use of such models in a clinical and developmental context is the reliability of predictions. Here we detail how the approach of using "models of models" has the potential to impact cancer treatment and drug development. We describe the iterative refinement process that leverages the flexibility of experimental systems to generate highly dimensional data, which can be used to train and validate computational model parameters and improve model predictions. In this way, highly optimized computational models with robust predictive capacity can be generated. Such models open up a number of opportunities for cancer drug treatment and development, from enhancing the design of experimental studies, reducing costs, and improving animal welfare, to increasing the translational value of results generated.

Cancer Precision Medicine: Why More Is More and DNA Is Not Enough.

Every tumour is different. They arise in patients with different genomes, from cells with different epigenetic modifications, and by random processes affecting the genome and/or epigenome of a somatic cell, allowing it to escape the usual controls on its growth. Tumours and patients therefore often respond very differently to the drugs they receive. Cancer precision medicine aims to characterise the tumour (and often also the patient) to be able to predict, with high accuracy, its response to different treatments, with options ranging from the selective characterisation of a few genomic variants considered particularly important to predict the response of the tumour to specific drugs, to deep genome analysis of both tumour and patient, combined with deep transcriptome analysis of the tumour. Here, we compare the expected results of carrying out such analyses at different levels, from different size panels to a comprehensive analysis incorporating both patient and tumour at the DNA and RNA levels. In doing so, we illustrate the additional power gained by this unusually deep analysis strategy, a potential basis for a future precision medicine first strategy in cancer drug therapy. However, this is only a step along the way of increasingly detailed molecular characterisation, which in our view will, in the future, introduce additional molecular characterisation techniques, including systematic analysis of proteins and protein modification states and different types of metabolites in the tumour, systematic analysis of circulating tumour cells and nucleic acids, the use of spatially resolved analysis techniques to address the problem of tumour heterogeneity as well as the deep analyses of the immune system of the patient to, e.g., predict the response of the patient to different types of immunotherapy. Such analyses will generate data sets of even greater complexity, requiring mechanistic modelling approaches to capture enough of the complex situation in the real patient to be able to accurately predict his/her responses to all available therapies.

The human gut virome: form and function.

Advances in next-generation sequencing technologies and the application of metagenomic approaches have fuelled an exponential increase in our understanding of the human gut microbiome. These approaches are now also illuminating features of the diverse and abundant collection of viruses (termed the virome) subsisting with the microbial ecosystems residing within the human holobiont. Here, we focus on the current and emerging knowledge of the human gut virome, in particular on viruses infecting bacteria (bacteriophage or phage), which are a dominant component of this viral community. We summarise current insights regarding the form and function of this 'human gut phageome' and highlight promising avenues for future research. In doing so, we discuss the potential for phage to drive ecological functioning and evolutionary change within this important microbial ecosystem, their contribution to modulation of host-microbiome interactions and stability of the community as a whole, as well as the potential role of the phageome in human health and disease. We also consider the emerging concepts of a 'core healthy gut phageome' and the putative existence of 'viral enterotypes' and 'viral dysbiosis'.

Predictive Modeling of Drug Treatment in the Area of Personalized Medicine.

Despite a growing body of knowledge on the mechanisms underlying the onset and progression of cancer, treatment success rates in oncology are at best modest. Current approaches use statistical methods that fail to embrace the inherent and expansive complexity of the tumor/patient/drug interaction. Computational modeling, in particular mechanistic modeling, has the power to resolve this complexity. Using fundamental knowledge on the interactions occurring between the components of a complex biological system, large-scale in silico models with predictive capabilities can be generated. Here, we describe how mechanistic virtual patient models, based on systematic molecular characterization of patients and their diseases, have the potential to shift the theranostic paradigm for oncology, both in the fields of personalized medicine and targeted drug development. In particular, we highlight the mechanistic modeling platform ModCell™ for individualized prediction of patient responses to treatment, emphasizing modeling techniques and avenues of application.

The human gut virome: a multifaceted majority.

Here, we outline our current understanding of the human gut virome, in particular the phage component of this ecosystem, highlighting progress, and challenges in viral discovery in this arena. We reveal how developments in high-throughput sequencing technologies and associated data analysis methodologies are helping to illuminate this abundant 'biological dark matter.' Current evidence suggests that the human gut virome is a highly individual but temporally stable collective, dominated by phages exhibiting a temperate lifestyle. This viral community also appears to encode a surprisingly rich functional repertoire that confers a range of attributes to their bacterial hosts, ranging from bacterial virulence and pathogenesis to maintaining host-microbiome stability and community resilience. Despite the significant advances in our understanding of the gut virome in recent years, it is clear that we remain in a period of discovery and revelation, as new methods and technologies begin to provide deeper understanding of the inherent ecological characteristics of this viral ecosystem. As our understanding increases, the nature of the multi-partite interactions occurring between host and microbiome will become clearer, helping us to more rationally define the concepts and principles that will underpin approaches to using human gut virome components for medical or biotechnological applications.

Network and systems biology: essential steps in virtualising drug discovery and development.

The biological processes that keep us healthy or cause disease, as well as the mechanisms of action of possible drugs are inherently complex. In the face of this complexity, attempts at discovering new drugs to treat diseases have alternated between trial-and-error (typically on experimental systems) and grand simplification, usually based on much too little information. We now have the chance to combine these strategies through establishment of 'virtual patient' models, centred on a detailed molecular characterisation of thousands or even, in the future, millions of patients. In doing so, we lay the foundations for truly personalised therapy, as well as a far-reaching virtualisation of drug discovery and development in oncology and other areas of medicine.

Disruption of Escherichia coli Nissle 1917 K5 capsule biosynthesis, through loss of distinct kfi genes, modulates interaction with intestinal epithelial cells and impact on cell health.

Escherichia coli Nissle 1917 (EcN) is among the best characterised probiotics, with a proven clinical impact in a range of conditions. Despite this, the mechanisms underlying these "probiotic effects" are not clearly defined. Here we applied random transposon mutagenesis to identify genes relevant to the interaction of EcN with intestinal epithelial cells. This demonstrated mutants disrupted in the kfiB gene, of the K5 capsule biosynthesis cluster, to be significantly enhanced in attachment to Caco-2 cells. However, this phenotype was distinct from that previously reported for EcN K5 deficient mutants (kfiC null mutants), prompting us to explore further the role of kfiB in EcN:Caco-2 interaction. Isogenic mutants with deletions in kfiB (EcNΔkfiB), or the more extensively characterised K5 capsule biosynthesis gene kfiC (EcNΔkfiC), were both shown to be capsule deficient, but displayed divergent phenotypes with regard to impact on Caco-2 cells. Compared with EcNΔkfiC and the EcN wild-type, EcNΔkfiB exhibited significantly greater attachment to Caco-2 cells, as well as apoptotic and cytotoxic effects. In contrast, EcNΔkfiC was comparable to the wild-type in these assays, but was shown to induce significantly greater COX-2 expression in Caco-2 cells. Distinct differences were also apparent in the pervading cell morphology and cellular aggregation between mutants. Overall, these observations reinforce the importance of the EcN K5 capsule in host-EcN interactions, but demonstrate that loss of distinct genes in the K5 pathway can modulate the impact of EcN on epithelial cell health.

Identification of aminoglycoside and ß-lactam resistance genes from within an infant gut functional metagenomic library.

The infant gut microbiota develops rapidly during the first 2 years of life, acquiring microorganisms from diverse sources. During this time, significant opportunities exist for the infant to acquire antibiotic resistant bacteria, which can become established and constitute the infant gut resistome. With increased antibiotic resistance limiting our ability to treat bacterial infections, investigations into resistance reservoirs are highly pertinent. This study aimed to explore the nascent resistome in antibiotically-naïve infant gut microbiomes, using a combination of metagenomic approaches. Faecal samples from 22 six-month-old infants without previous antibiotic exposure were used to construct a pooled metagenomic library, which was functionally screened for ampicillin and gentamicin resistance. Our library of ∼220Mb contained 0.45 ampicillin resistant hits/Mb and 0.059 gentamicin resistant hits/Mb. PCR-based analysis of fosmid clones and uncloned metagenomic DNA, revealed a diverse and abundant aminoglycoside and ß-lactam resistance reservoir within the infant gut, with resistance determinants exhibiting homology to those found in common gut inhabitants, including Escherichia coli, Enterococcus sp., and Clostridium difficile, as well as to genes from cryptic environmental bacteria. Notably, the genes identified differed from those revealed when a sequence-driven PCR-based screen of metagenomic DNA was employed. Carriage of these antibiotic resistance determinants conferred substantial, but varied (2-512x), increases in antibiotic resistance to their bacterial host. These data provide insights into the infant gut resistome, revealing the presence of a varied aminoglycoside and ß-lactam resistance reservoir even in the absence of selective pressure, confirming the infant resistome establishes early in life, perhaps even at birth.

Personalized medicine approaches for colon cancer driven by genomics and systems biology: OncoTrack.

The post-genomic era promises to pave the way to a personalized understanding of disease processes, with technological and analytical advances helping to solve some of the world's health challenges. Despite extraordinary progress in our understanding of cancer pathogenesis, the disease remains one of the world's major medical problems. New therapies and diagnostic procedures to guide their clinical application are urgently required. OncoTrack, a consortium between industry and academia, supported by the Innovative Medicines Initiative, signifies a new era in personalized medicine, which synthesizes current technological advances in omics techniques, systems biology approaches, and mathematical modeling. A truly personalized molecular imprint of the tumor micro-environment and subsequent diagnostic and therapeutic insight is gained, with the ultimate goal of matching the "right" patient to the "right" drug and identifying predictive biomarkers for clinical application. This comprehensive mapping of the colon cancer molecular landscape in tandem with crucial, clinical functional annotation for systems biology analysis provides unprecedented insight and predictive power for colon cancer management. Overall, we show that major biotechnological developments in tandem with changes in clinical thinking have laid the foundations for the OncoTrack approach and the future clinical application of a truly personalized approach to colon cancer theranostics.

Effects of selective digestive decontamination (SDD) on the gut resistome.

OBJECTIVES: Selective digestive decontamination (SDD) is an infection prevention measure for critically ill patients in intensive care units (ICUs) that aims to eradicate opportunistic pathogens from the oropharynx and intestines, while sparing the anaerobic flora, by the application of non-absorbable antibiotics. Selection for antibiotic-resistant bacteria is still a major concern for SDD. We therefore studied the impact of SDD on the reservoir of antibiotic resistance genes (i.e. the resistome) by culture-independent approaches. METHODS: We evaluated the impact of SDD on the gut microbiota and resistome in a single ICU patient during and after an ICU stay by several metagenomic approaches. We also determined by quantitative PCR the relative abundance of two common aminoglycoside resistance genes in longitudinally collected samples from 12 additional ICU patients who received SDD. RESULTS: The patient microbiota was highly dynamic during the hospital stay. The abundance of antibiotic resistance genes more than doubled during SDD use, mainly due to a 6.7-fold increase in aminoglycoside resistance genes, in particular aph(2″)-Ib and an aadE-like gene. We show that aph(2″)-Ib is harboured by anaerobic gut commensals and is associated with mobile genetic elements. In longitudinal samples of 12 ICU patients, the dynamics of these two genes ranged from a ∼10(4) fold increase to a ∼10(-10) fold decrease in relative abundance during SDD. CONCLUSIONS: ICU hospitalization and the simultaneous application of SDD has large, but highly individualized, effects on the gut resistome of ICU patients. Selection for transferable antibiotic resistance genes in anaerobic commensal bacteria could impact the risk of transfer of antibiotic resistance genes to opportunistic pathogens.